| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2000;69:1817-1821
© 2000 The Society of Thoracic Surgeons
a Division of Cardiothoracic Surgery, Department of Surgery, School of Medicine of the University of California, San Francisco and the San Francisco Veterans Affairs Medical Center, San Francisco, California, USA
Address reprint requests to Dr Ratcliffe, VAMC Surgery 112D, San Francisco Veterans Affairs Medical Center, 4150 Clement St, San Francisco, CA 94121
e-mail: ratcliffe.mark{at}sanfrancisco.va.gov
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and methodsResultsCommentReferences |
|---|
Methods. A Java-based applet was developed and deployed on a hospital intranet. Briefly, the nurse entered patient process data using a point and click interface. The applet generated a nursing note, and process data were saved in a Microsoft Access database. In 10 patients, this method was validated by comparison with a retrospective chart review. In 45 consecutive patients, weekly control charts were generated from the data. When aberrations from the pathway occurred, feedback was initiated to restore the goals of the critical pathway.
Results. The intranet process data collection method was verified by a manual chart review with 98% sensitivity. The control charts for time to extubation, intensive care unit stay, and hospital stay showed a deviation from critical pathway goals after the first 20 patients. Feedback modulation was associated with a return to critical pathway goals.
Conclusions. Java-based applets are inexpensive and can collect accurate postoperative process data, identify critical pathway deviations, and allow timely feedback of process data.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentReferences |
|---|
Accurate process data are difficult and expensive to obtain [8–11]. Nurses who perform hospital-based utilization review or who act as case managers for certain subsets of patients are knowledgeable but expensive to employ. Data obtained from hospital and medical records data systems are often inaccurate, temporally inefficient, and not focused on daily patient events [12].
An optimal data collection system should be inexpensive, reliable, and allow the almost immediate availability of data so that variances from critical pathway goals can be rapidly identified and feedback of information provided to responsible physicians. We designed a Java-based intranet program (applet) that integrates progress note generation with postoperative data collection and tested the hypothesis that the Java-based system is inexpensive, accurate, and able to provide rapid feedback of process data so that critical pathway goals are maintained.
|
Material and methods |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentReferences |
|---|
|
User interface
Briefly, the nurse opens the "Cardiac Surgery Critical Pathway" applet using a standard web browser. The first page identifies the user and the patient for secure access (Fig 2A). The second page displays 11 categories of variables (location, activity, medication, etc), with a total of 75 true/false variables (Fig 2B). The nurse uses a point and click interface to select the appropriate entry for each variable. The third and final page displays an automatically generated text progress note (Fig 2C), which can be edited by the nurse before the note is automatically entered into the patient’s electronic medical record. Data were entered at the end of each 12-hour nursing shift.
|
Critical pathway
A multidisciplinary team developed a critical pathway for coronary artery bypass grafting. At the time of data collection, our critical pathway goals included extubation at 6 hours, intensive care unit discharge at 2 days, and a 6-day hospital stay.
Validation of the system
The reliability of the system was validated in 10 consecutive patients undergoing coronary artery bypass grafting. A manual retrospective chart review was performed and 50 process variables collected by both the intranet system and chart review were compared. After completion of the pilot study, all nurses who used the intranet system were asked to rate the ease of use of the system on a five-point Leikert scale (very easy to use, easy to use, OK to use, difficult to use, very difficult to use).
Feedback modulation
After validation, the intranet system was used to prospectively collect process data in 45 consecutive patients undergoing coronary artery bypass grafting. Three measures used in the critical pathway—time to extubation, intensive care unit stay, and duration of hospitalization—were evaluated weekly, using control charts (Proc Shewart, SAS System for Windows, version 6.12, The SAS Institute, Cary, NC) [13]. Control limits were set at three standard deviations from the mean. Adherence to the goals of the production process could be appreciated and aberrations noticed immediately as points exceeding the upper control limit. When systematic aberrations were noted, an attempt was made to identify the sources of the aberration and feedback was given to the entire cardiac care team.
|
Results |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentReferences |
|---|
The average time for a nurse to enter the required electronic progress note before the initiation of the intranet-based system is 7 minutes. Using the intranet system, the average time for generation of a nursing progress note is 4 minutes. There was no evidence of a significant learning curve, as the time for data entry stayed the same throughout the course of the study. The 24-hour technical support number available to the nurses was never used. There were no hardware or software failures during the testing.
Validation of the system
Ten consecutive patients undergoing coronary artery bypass grafting were entered into the study. Validation of the intranet system was carried out by comparing data on the intranet system with information obtained on manual review of the hospital medical record (Table 1). Among a total of 50 variables assessed, only a single point of disagreement between the intranet system and chart review was identified. This was accounted for by a 12-hour interval between extubation and the nurse’s progress note. Overall, data from the intranet system was confirmed by the chart review with an accuracy of 98%. The total time for data analysis with the intranet system was 36 minutes, compared to 5 hours 10 minutes for the retrospective chart review. A total of 13 nurses entered data on the system. All of them rated the system "very easy to use."
|
Control charts for time to extubation, duration of stay in the intensive care unit, and duration of hospitalization are presented in Figure 3. Persistent aberrations in all three process measurements were identified after approximately 20 patients. Aberrations were associated with a change in the resident staff and although no single causative factor was identified, multiple factors were identified as probably contributory. Specifically, the time when propofol was stopped, the time when chest tubes were discontinued, and the initiation and aggressiveness of diuresis were all thought to be contributory. Potential sources were discussed with the entire cardiac care team and the guidelines of the pathway reinforced. With this feedback modulation, performance of all three measures returned to critical pathway goals within 10 patients of the aberration. By independent sample t tests, the duration of intubation (p = 0.001), stay in the intensive care unit (p < 0.001), and hospitalization (p < 0.001) were significantly longer during the period of systemic aberration (patients 22 to 31) than periods during which the pathway goals were regularly met. The duration of inotropic support was longer during the period of aberration to a degree that approached statistical significance (p = 0.06).
|
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentReferences |
|---|
The collection of accurate patient process data and timely review by responsible physicians is necessary if a program of continuous quality improvement is to be implemented and if critical pathways are to achieve their optimal effectiveness [5–7]. However, accurate process data are difficult and expensive to obtain. Data obtained from hospital and medical records data systems are often inaccurate, temporally inefficient, and not focused on daily patient events [12]. Traditionally, outcome data in cardiac operations have been based on retrospective reviews or multiinstitutional databases, such as those maintained by the Health Care Financing Administration, the Department of Veterans Affairs, and The Society of Thoracic Surgeons [7, 12, 14–16]. Among the difficulties with optimizing outcome-based cardiac surgical care are the expense and problems with these large databases, including lack of standardization, cost of dedicated data entry/delivery personnel, and lack of timely feedback [12]. We do not mean to minimize the importance of the large multiinstitutional cardiac surgery databases. However, those systems were not designed to provide rapid feedback of process data, which is necessary if individual practitioners are to effectively manage cardiac surgery pathways.
There has been an explosion in internet-based computer programs, programming languages, and technologies in recent years. The Java programming language, although initially developed by Sun Microsystems to program electrical appliances, has been recognized as an ideal tool for distributed programming over the internet. Programs written in Java are object oriented, machine independent, and compact. Spurred by the desire to take product orders over the internet, Java classes that implement database connections to standard personal computer and workstation-based database programs have been developed. These events promise to revolutionize the collection and distribution of data over local (intranet) and worldwide (internet) networks, and are very well suited for the development of a system for collecting process data that would be inexpensive to develop, straightforward to apply, and reliable.
Because programs written in Java are compact and machine independent, client computers do not need high speed or disk space and therefore, systems may be inexpensive. When amortized over 5 years, with an estimated 200 patients per year, the cost of our system comes to $8.64 per patient for data collection. In this study, all of the programming and data analysis was performed by one of the authors (MBR). It is admittedly unusual for a clinician to have these skills and it is more likely that projects of this would have to be completed by outside contractors. It is also true that the Veterans Administration with its intranet and electronic medical record provides an infrastructure that is not present in most hospital environments. The actual cost of system implementation may be higher than estimated here.
One of the factors that facilitated implementation of this system at the San Francisco Veterans Affairs Medical Center was the existing electronic medical record, which requires that nurses enter a computerized progress note. Thus, the nurses were accustomed to using a computer-based charting system. Integration of the data collection with the generation of a nursing progress note clearly provides the motivation for the bedside nurse to use the system. Moreover, we were able to couple our data collection system with the requirement for an electronic progress note efficiently, so that no additional effort was required of the nurses. In fact, the intranet-based system decreased by 50% the amount of time the nurses spent charting their progress note.
Control charts continuously track process variables with their mean and upper and lower control limits [13]. Their application makes the case for using an industrial quality management science in health care organization [17]. Automatic data collection and subsequent generation of weekly control charts was a cost effective method for providing timely feedback. The benefits of timely feedback with the control chart method are strongly suggested by this study.
Although this study represents a preliminary endeavor in this area, we are encouraged by our findings. To date we found only one example of a limited data collection system for Trauma using the internet [18]. We realize that this specific system is fairly simple and will not be universally applicable, but it does highlight the potential for using internet-based technologies to facilitate effective and economical healthcare delivery. We anticipate that the use of computer technology of this sort in health information management will become increasingly common and sophisticated.
|
References |
|---|
|
TopAbstractIntroductionMaterial and methodsResultsCommentReferences |
|---|
Related Article
Ann. Thorac. Surg. 2000 69: 1821-1822.
This article has been cited by other articles:
|
|
 |
|
  J. Thor, J. Lundberg, J. Ask, J. Olsson, C. Carli, K. P. Harenstam, and M. Brommels Application of statistical process control in healthcare improvement: systematic review Qual. Saf. Health Care, October 1, 2007; 16(5): 387 - 399. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 T. Aberg and J. Hentschel Improved total quality by monitoring of a cardiothoracic unit. Medical, administrative and economic data followed for 9 years Interactive CardioVascular and Thoracic Surgery, March 1, 2004; 3(1): 33 - 40. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
